Skip to main content

Table 3 Quality of evidence and summary of findings table

From: School-based sexual health education interventions to prevent STI/HIV in sub-Saharan Africa: a systematic review and meta-analysis

  Quality assessment № of participants Effect Quality
Outcome (follow up period) № of studies Study design Risk of biasa Inconsistencyb Indirectnessc Imprecision Publication or reporting bias School-based sexual health education Nothing or usual curriculum Relative (95 % CI) Absolute (95 % CI)
Herpes Simplex Virus- 2 (54 months) 3 randomised trials not serious not serious not serious not seriousd none 746/6146 (12.1 %) 738/6127 (12.0 %) OR 1.07 (0.94 to 1.23) 7 more per 1,000 (from 6 fewer to 24 more) HIGH
*Condom use (less than six months) 4 randomised trials not serious not serious not serious seriouse none 332/741 (44.8 %) 295/720 (41.0 %) OR 1.62 (1.03 to 2.55) 120 more per 1,000 (from 7 more to 229 more) MODERATE
*Condom use (less than six months) 4 quasi-experimental trials not serious not serious not serious Seriousf None 90/226 (39.8 %) 63/197 (32.0 %) OR 2.88 (1.41 to 5.90) 255 more per 1,000 (from 79 more to 415 more) LOW
*Condom use (six to 10 months) 6 randomised trials not serious not serious not serious not seriousg none 485/1238 (39.2 %) 494/1494 (33.1 %) OR 1.40 (1.16 to 1.68) 78 more per 1,000 (from 34 more to 123 more) HIGH
Condom use (six to 10 months 2 quasi-experimental trials not serious not serious not serious serioush publication bias strongly suspectedi 62/146 (42.5 %) 55/131 (42.0 %) OR 1.05 (0.65 to 1.71) 12 more per 1,000 (from 100 fewer to 133 more) VERY LOW
*Condom use (more than 10 months) 7 randomised trials not serious not serious not serious not seriousj none 2955/8106 (36.5 %) 2678/8868 (30.2 %) OR 1.22 (0.99 to 1.50) 43 more per 1,000 (from 2 fewer to 92 more) HIGH
Condom use (more than 10 months) 4 quasi-experimental trials not serious not serious not serious Seriousk none 1442/2205 (65.4 %) 1647/2649 (62.2 %) OR 1.18 (0.92 to 1.52) 38 more per 1,000 (from 20 fewer to 92 more) VERY LOW
  1. CI Confidence interval, OR Odds ratio, *outcomes with statistically significant positive results
  2. aQuality was assessed as ‘not serious’ for all outcomes because majority of the studies included were of ‘low’ to ‘moderate’ risk of bias
  3. bI2 were all below 75 % and therefore, quality was assessed as ‘not serious’ for all the outcomes
  4. cQuality was assessed as ‘not serious’ for all outcomes because the interventions were fairly similar, participants were adolescents or young adults and all assessed similar outcomes (condom use and laboratory test of Herpes Simplex Virus-2)
  5. dRelative risk increase or decrease is less than 25 % (95 % CI of 0.97−1.23)
  6. eRelative risk increase or decrease is greater than 25 % (95 % CI of 1.03 to 2.55)
  7. fRelative risk increase or decrease is greater than 25 % (95 % CI of 1.41 to 5.90)
  8. gRelative risk increase or decrease is very close to 25 % confidence interval (95 % CI of 1.16 to 1.68)
  9. hRelative risk increase or decrease is greater than 25 % (95 % CI of 0.65 to 1.75)
  10. iFunnel plot was asymmetrical
  11. jRelative risk increase or decrease is very close to 25 % (95 % CI of 0.99 to 1.50)
  12. kRelative risk increase or decrease is greater than 25 % (95 % CI of 0.92 to 1.52)